Health Care Industry Consolidation

Statement before the Committee on Ways and Means Health SubcommitteeU.S. House of Representatives

by

Martin GaynorE.J. Barone Professor of Economics and Health Policy

Heinz CollegeCarnegie Mellon University

Washington, D.C.September 9, 2011

1 Introduction

Health care is a very large and important industry. Hospital and physician services are alarge part of the U.S. economy. In 2009, hospital care alone accounted for 5.4% of GDP– roughly twice the size of automobile manufacturing, agriculture, or mining, and largerthan all manufacturing sectors except food and beverage and tobacco products, which isapproximately the same size. Physician services comprise 3.6% of GDP (Martin et al.,2011). The net cost of health insurance – current year premiums minus current yearmedical benefits paid – was 1% of GDP in 2009. The share of the economy accountedfor by these sectors has risen dramatically over the last 30 years. In 1980, hospitals andphysicians accounted for 3.6% and 1.7% of U.S. GDP, respectively, while the net cost ofhealth insurance in 1980 was 0.34% (Martin et al., 2011).

Of course, health care is important not only because of its size. Health care servicescan save lives or dramatically affect the quality of life, thereby substantially improvingwellbeing and productivity.

As a consequence, the functioning of the health care sector is vitally important. Awell functioning health care sector is an asset to the economy and improves quality oflife for the citizenry. By the same token, problems in the health care sector act as a dragon the economy and impose a burden on individuals.

As documented below, there has been a tremendous amount of consolidation amonghealth care providers. Consolidation has also been occuring among health insurers. Con-solidation can bring efficiencies – it can reduce inefficient duplication of services, allowfirms to combine to achieve efficient size, or facilitate investment in quality or efficiencyimprovements. On the other hand, consolidation can enhance the market power and leadto increased prices or reduced quality.

The research evidence shows that providers in more concentrated markets charge

1

higher prices to private payers, without accompanying gains in efficiency or quality. Fur-ther, the burden of higher provider prices falls on individuals, not insurers or employers.Even though individuals with private employer provided health insurance pay a smallportion of provider fees directly out of their own pockets, they end up paying for in-creased prices in the end. Insurers facing higher provider prices increase their premiumsto employers. Employers then pass those increased premiums on to their workers, eitherin the form of lower wages (or smaller wage increases) or reduced benefits (greater pre-mium sharing or less extensive coverage, including the loss of coverage). There is lessresearch evidence on the impacts of consolidation in health insurance markets, but thatevidence shows insurers in more concentrated markets charge higher premiums to largeemployers, and pay lower prices to providers.

This represents a real, and serious problem. The U.S. health care system depends onprivate markets to deliver health care and for a large part of the financing of care. Ifthese markets are not functioning well due to the exercise of market power this createsproblems for the entire system.

2 Consolidation

While not a new phenomenon, there has been substantial consolidation in the health careindustry in recent years. There was a large amount of consolidation in hospital marketsin the 1990s. Consolidation activity slowed down starting around 2002, but has turnedup in the past couple of years.

Figure 1 presents information on hospital merger and acquisition activity from 1998through the first two quarters of 2011. As can be seen, there were a large number ofmergers and acquisitions in the 1990s, with the pace slowing down in the early 2000s,and some pick up in activity in the past couple of years.

Some additional information is contained in Figure 2, which charts the number ofhospitals that are members of systems from 2000 to 2009. Members of hospital systemsare often jointly owned and operated, so can be considered to be one firm. As can beseen, membership has been increasing substantially over time, thereby diminishing thenumber of independent hospitals.

The Herfindahl-Hirschmann Index (HHI) is a measure of how concentrated shares arewithin a market, and is widely employed as a measure of market structure. The HHI isthe sum of squared market shares in the market. It increases as market shares are moreconcentrated among a small number of firms. It reaches its maximum value of 10,000 fora monopoly (the square of the monopolist’s market share of 100 percent), and reaches aminimum value when the market is equally divided.

Table 1 presents numbers for the population-weighted, Herfindahl-Hirschmann Index

2

for hospitals for selected years from 1987 to 2006.1 Two things are clear from this table.U.S. hospital markets are highly concentrated and have become even more concentratedover time. From the table it is easily seen that hospital markets have become significantlymore concentrated. In 1987, the mean HHI was 2,340 and by 2006 the HHI was 3,161 – anincrease of over 900 points. In 1992, the mean hospital concentration levels (2,440) were(barely) below the recently updated Federal merger guidelines’ (Federal Trade Commis-sion and Department of Justice, 1992) cut-off point for classifying a market as “HighlyConcentrated” (HHI ≥ 2,500), but by 2006 the mean concentration level (3,261) rose towell above this threshold. Town et al. (2006) note that mergers and acquisitions are theprimary reason for the increase in hospital concentration over this period.

To provide some context for these numbers, a market with 5 equally sized firms wouldhave a HHI of 2,000, 4 equally sized firms would have an HHI of 2,500 and a market with3 equally sized firms would have an HHI of 3,333. That is, the increase in the HHI overthe last 20 years is the equivalent of moving from approximately 5 equally sized firms toa market with about 3 equally sized firms.

While hospital markets are highly concentrated on average, there is also wide variationin concentration. Figure 3 shows a scatterplot of the MSA level market concentration in1990 and in 2006. This figure displays two phenomena. First, it shows the distribution ofHHIs across MSAs. Most MSAs are “Highly Concentrated.” In 2006, of the 332 MSAsin the U.S., 250 had HHIs greater than 2,500. Second, it is clear from Figure 3 thatthe increase in hospital concentration was a broad phenomenon – the vast majority ofMSAs became more concentrated over this period. Particularly striking is the number ofmoderately concentrated MSAs in 1990 that by 2006 had become highly concentrated.

An obvious question is why this wave of hospital consolidation occurred. Fuchs (2007)and others point to the rise of managed care as the principal factor driving this massiveconsolidation. A cursory glance at Figure 4 suggests this causal explanation. The ideais that the rise of HMOs introduced aggressive price negotiations between hospitals andhealth plans, thereby giving hospitals a strong incentive to acquire bargaining powerthrough consolidation. The rise of HMOs during the 1990s is widely credited with sig-nificantly reducing health care cost growth, primarily through tough price negotiations(see, e.g., Cutler et al., 2000).

Early suggestive evidence is provided by Chernew (1995), who finds that in the 1980sthere is a relationship between HMO penetration and the number of hospitals operatingin the market. Dranove et al. (2002) examine data from the 1981 to 1994 and finda correlation between metropolitan area HMO penetration in 1994 and the change inmarket structure. However, Town et al. (2007) examine the change in hospital market

1The table contains population weighted, averages for all but the largest Metropolitan StatisticalAreas (MSA)(based on admissions). The sample is limited to those MSAs with a population less than3 million in 1990 because it is likely there are multiple hospital markets in MSAs larger than 3 millionpopulation .

3

structure and the change in HMO penetration and find little correlation. Some havesuggested that it wasn’t the realization of the rise of managed care, but the anticipation(which in some cases may have been in error) that led hospitals to consolidate. Workin progress by Town and Park (2011) provides support for this hypothesis. They findthat HMO exit, a measure of the exuberance of expectations regarding the demand formanaged care in a location, is correlated with hospital consolidation.

There have also been substantial changes in market structure in U.S. physician mar-kets. Liebhaber and Grossman (2007) report that the percent of physicians in solo or2 person practices declined from 40.7 percent in 1996-97 to 32.5 percent in 2004-05.Further, the proportion in practices of 3-5 physicians fell over the same period. Theproportion of physicians practicing in groups of 6 or more grew from 15.9 percent to 21.8percent. The number of physicians in other practice settings (primarily employed byothers) grew from 31.2 to 36.0 percent over this period. Since the number of physiciansper 1,000 persons has not really changed (∼2.5) since 1997 (National Center for HealthStatistics, 2011), this represents an increase in concentration.

There is no good systematic information on the structure of local physician mar-kets. Those markets, especially for specialized services, may be very concentrated, butthere is no information generally available at the national level. Schneider et al. (2008)constructed HHIs for physician organizations in California at the county level for 2001.They find the average county HHI for physician organizations was 4,430, implying a highdegree of concentration on average. They found that 17 percent of California countieshad a physician organization HHI below 1,800, 33 percent had an HHI between 1,800and 3,600, and 50 percent had an HHI above 3,600.

Tables 2, 3, and 4 provide information about health insurance market structuresfor the U.S. This information shows consistently high levels of concentration in healthinsurance markets.

Table 2 contains measures of HHI for HMO plus PPO markets in the U.S. from reportsfrom the American Medical Association (AMA). They show high levels of concentration(although lower than for hospitals). The numbers show insurance market concentrationdeclining somewhat over time (although not consistently). However, there are someconcerns about the accuracy of these numbers (see Capps, 2009; Dafny et al., 2011a).

Dafny (2010) and Dafny et al. (2011b), using data on the large employer segment ofthe insurance market, also show increasing concentration in health insurance markets.Dafny (2010, Figure 5) documents an increase in the percentage of markets with 1-4, 5-6,or 7-9 insurance carriers in the U.S. from 1998-2005, and a decrease in the percentage ofmarkets with 9-10 or more than 10 carriers. Dafny et al. (2011b) state that the meanHHI in their sample increased from 2,286 to 2,984 from 1998-2006, the median 4 firmconcentration ratio increased from 79 to 90 percent, and the mean number of carriersper market fell from 18.9 to 9.6. They show (Figure 1 in their paper) that 78 percentof the markets they study had increases in the HHI of 100 points or more from 2002

4

to 2006, and 53 percent experienced increases of 500 points or more. Table 3 has meanHHIs by year from the data used in those papers. These numbers indicate that the largeemployer segment of the health insurance market is concentrated and has grown more soover time. These numbers are roughly similar in magnitude to those calculated by theAMA. However they show concentration increasing over time (by about 400 points from2004-2008), while the AMA numbers exhibit a slight decrease over time.

A recent report by the U.S. Government Accountability Office (Government Account-ability Office, 2009) compiled information on the market structure of the small grouphealth insurance market in the U.S. Table 4 reproduces numbers from that report. Ascan be seen, those markets appear to be fairly heavily concentrated, and increasing inconcentration. A recent paper by Schneider et al. (2008) utilizes a unique data source forCalifornia to construct HHIs for insurance plans at the county level for 2001. They findan average insurance HHI for California counties of 2,592. They report that 21 percentof counties have HHIs below 1,8002, 55 percent had HHIs between 1,800 and 3,600 and24 percent had HHIs above 3,600. The information from these various data sources seembroadly consistent.

It is worth noting that, based on the data in Table 3 the increase in concentration inhealth insurance markets appears to occur much later than the increase in concentrationin hospital markets. The major increase in hospital market concentration occurred in themid to late 1990s. Insurance market concentration, at least for large employers, startsincreasing in 2002.

In addition to “horizontal” consolidation between firms in the same industry (hospitalsand hospitals, physician practices and physician practices), there has been some “vertical”consolidation between hospitals and physician practices. Figure 5 shows the trends in theproportions of hospitals with various kinds of integration with physicians. As can be seen,various forms of integration between physicians and hospitals grew and peaked duringthe 1990s, and have declined since then, with the exception of physician employment byhospitals, which has increased substantially over this time period.

In addition to information on market structure, there is some information on trendsin prices and the contribution of health care prices to overall health care cost growth.Akosa Antwi et al. (2009) document a 100% increase in hospital prices in California from1999-2006, although they do not find market concentration to be a contributor to theincrease. Martin et al. (2011) decompose U.S. health spending growth into growth dueto prices versus growth due to nonprice factors (e.g., population, intensity of care). Theyfind that prices account for 60% of the increase in overall spending from 2008 to 2009.The proportion of health spending growth due to prices varies over time (see Exhibit 6in Martin et al., 2011), but has been growing steadily since 2001.

2The old Federal merger guidelines cutoff for considering a market highly concentrated (Federal TradeCommission and Department of Justice, 1992).

5

Some recent reports from state governments document growth or variation in healthcare prices. A report from the Massachusetts Attorney General’s office (MassachusettsAttorney General, 2010) finds that price increases caused most of the increases in healthcare spending in the state in recent years. The report also finds significant variation inprices and that the variation is uncorrelated with quality of care, but is correlated withmarket leverage. A report on Pennsylvania hospitals found substantial variation in pricesfor heart surgery, but no correlation of prices with quality (Pennsylvania Health CareCost Containment Council, 2007).

Overall, the statistics presented here paint a picture of health care markets that areconcentrated and becoming more so over time. There is also some evidence that pricesare rising faster than quantities, and that price variation isn’t related to quality but maybe due to market power.

3 Evidence on the Impacts of Consolidation

Most of the evidence on the impacts of consolidation come from hospital markets. Un-fortunately there are very few studies of competition in physician markets, mainly dueto lack of data.

3.1 Hospitals

3.1.1 Price

There has been a lot of research on the impact of hospital market consolidation on pricespaid by private payers. The overwhelming finding in the literature is that consolidationleads to higher prices (see Dranove and Satterthwaite, 2000; Gaynor and Vogt, 2000;Vogt and Town, 2006; Gaynor and Town, 2012, for reviews of the evidence).

Examining the distribution of realized hospital prices (for the privately insured) aloneis informative about the functioning of hospital markets. Ginsburg (2010) uses admin-istrative claims data for 8 geographic areas from 4 large private insurers to constructinpatient hospital prices. He finds that there is significant variation both within andacross regions in hospital prices. For example, San Francisco has the highest averagehospital prices in 2008, with prices equal to 210% of the Medicare reimbursement rate.The lowest rate is Miami-South Florida with mean prices that are 147% of Medicare rates– the mean price in San Francisco is 43% higher than Miami. Within San Francisco, theinterquartile range is 116% of the Medicare price. Of course, there are a number of possi-ble reasons for this variation. Cost, quality and demand differences will generally implyprice differences. However, it seems unlikely that there is enough variation across thosefactors to generate such wide variation in price.

6

There are a number of different methods that have been used to estimate the impactof hospital consolidation on prices. The most direct approach compares price increasesat merging hospitals with those at similar hospitals which did not merge (see Capps andDranove, 2004; Dafny, 2009; Haas-Wilson and Garmon, 2011; Krishnan, 2001; Spanget al., 2001; Sacher and Vita, 2001; Tenn, 2011; Thompson, 2011). The vast majority ofthese studies find price increases of at least 10 percent due to merger, with some estimatesof price increases due to merger of 40 percent or greater.

For example, Haas-Wilson and Garmon (2011) evaluate the Evanston Northwesternand Highland Park hospitals in the northern suburbs of Chicago. They find a priceincrease of 20 percent due to that merger. Tenn (2011) examines the merger of twohospitals in California: Summit and Alta Bates. He finds that prices at Summit hospitalincreased between 28 and 44 percent after the merger.

Another source of information on the impacts of hospital consolidation comes fromstudies which examine the impact of hospital market concentration (measured as theHHI) on price. These studies don’t examine the effects of mergers directly, but allow oneto calculate the expected impact of a merger based on its impact on market concentration.Vogt and Town (2006) calculate the average estimated impact of a merger of two equalsized hospitals in a five hospital market (a “5 to 4” merger).3 They find that such amerger is estimated to increase prices by 5 percent.

Last, a few research papers have estimated the impacts of hospital mergers usingsimulation. These papers estimate models of hospital competition, then use the estimatedparameters of those models to simulate the impacts of mergers (Town and Vistnes, 2001;Capps et al., 2003; Gaynor and Vogt, 2003; Brand et al., 2011). These papers findestimated impacts of mergers ranging from 5 to 53 percent increases in price. Town andVistnes (2001) examine mergers among hospitals in Los Angeles and Orange Counties,California, where there are more than 120 hospitals between the two counties. They findthat many of the mergers they examine would result in price increases of 5 percent orgreater, in spite of the large number of hospitals in these counties. Capps et al. (2003)examine a 3 hospital merger in the southern suburbs of San Diego County, California, andfind a price increase due to the merger of over 10 percent. Gaynor and Vogt (2003) findthat a three-to-two hospital merger in San Luis Obispo, California (which was attempted,but blocked by the FTC) would have raised prices by over 50 percent. Brand et al. (2011)consider the recent proposed acquisition of Prince William hospital in Manassas, Virginiaby Inova health system in Northern Virginia. They estimate that the acquisition wouldhave led to price increases at Prince William hospital of anywhere from 19 to 33 percent.

Overall, these studies consistently show that hospital consolidation raises prices, andby nontrivial amounts. Consolidated hospitals that are able to charge higher prices dueto enhanced market power are able to do so on an ongoing basis, making this a permanent

3This results in an 800 point increase in the HHI, from 2000 to 2,800. As reported in 1, the averageHHI rose by about this amount from 1997 to 2002, albeit from a higher base.

7

rather than a transitory problem.

3.1.2 Not-for-Profit Firm Behavior

The hospital sector is characterized by the fact that there is a mixture of firms withdifferent ownership types. Not-for-profits are the most common, but there are substantialnumbers of for-profit hospitals and public hospitals. One question that is relevant in thissetting is whether not-for-profit hospitals behave any differently with regard to theirpricing behavior.

A number of studies (e.g., Keeler et al., 1999; Simpson and Shin, 1997; Dranove andLudwick, 1999; Capps et al., 2003; Gaynor and Vogt, 2003) have addressed the issue ofnot-for-profit/for-profit differences in competitive conduct. Those studies do not find anysignificant differences in pricing behavior. In particular, the effects of consolidation onpricing do not appear to differ depending on whether a hospital is not-for-profit.

A recent study by Capps et al. (2010) examines whether not-for-profit hospitals aremore likely than for-profit hospitals to offer more charity care or unprofitable services inresponse to an increase in market power. The implication is, that if there were such adifference, not-for-profits would be spending their profits from market power on sociallybeneficial activities. Capps et al. examine 7 years of data on California hospitals and findno evidence of any such differences – not-for-profits do not engage in any more sociallybeneficial activities than do for-profits when they possess market power.

3.1.3 Costs

It is clear that mergers can result in efficiencies because of economies of scale, increasedpurchasing power, the ability to consolidate services, or the the transfer of managerialtechniques and skill to the acquired hospital. However, mergers also have the potentialto increase costs. Larger systems imply larger bureaucracies. In addition, hospital costsare not necessarily exogenous to market structure. Hospitals that are able to bargain forhigher prices may have the incentive to use the resulting profits for the benefit of physi-cians and hospital executives (e.g., through capital expenditures that benefit physiciansor increases in executive compensation or perks). This is particularly likely if there isno residual claimant (as is the case for not-for-profit organizations) or monitoring by theresidual claimant is costly. Thus, the analysis of cost impacts is central to understandingthe impact of hospital mergers. The evidence presented above suggests that, on average,hospital mergers result in increases in price. Consequently if there are significant costreductions associated with mergers they are not passed onto the purchasers of hospitalservices in the form of lower prices.

A few studies do directly examine the impact of hospital mergers on costs. Dranoveand Lindrooth (2003) examine mergers of previously independent hospitals that consoli-

8

date financial reporting and operate under a single license post-merger. They find that,on average, these hospitals experience post-merger cost decreases of 14 percent. Systemmergers in which the hospitals were not as fully integrated (as measured by the use ofmultiple licenses) did not realize cost savings. These findings suggest that integrationof merging hospitals is necessary to achieve meaningful efficiencies. A recent study byHarrison (2010) finds that immediately following a merger costs declined, but eventuallyrose to pre-merger levels. This finding is difficult to reconcile with the view that mergersrequire significant upfront costs but have benefits accrue in later years. The circum-stances in which mergers are most likely to result in meaningful cost decreases are thosein which the merging facilities operate as a more fully integrated entity. To be clear,however, the presence of any cost savings does not mean that they are necessarily passedon to consumers.

3.1.4 Quality

A number of research studies have examined the impacts of hospital consolidation onvarious measures of quality, although the most commonly used measure of quality ismortality (adjusted for patient severity of illness). The results in this literature aremixed, although the results are strongest for markets with regulated prices (see Gaynor,2006; Vogt and Town, 2006; Gaynor and Town, 2012, for surveys).

A number of studies have examined the impact of market concentration on patientmortality for Medicare patients. There are a variety of findings, but the strongest studiesfind that market concentration significantly increases mortality (Kessler and McClellan,2000; Kessler and Geppert, 2005).

Kessler and McClellan find that risk-adjusted one year mortality for Medicare heartattack (acute myocardial infarction, or AMI) patients is significantly higher in moreconcentrated markets. In particular, patients in the most concentrated markets hadmortality probabilities 1.46 points higher than those in the least concentrated markets(this constitutes a 4.4% difference) as of 1991. This is an extremely large difference –it amounts to over 2,000 fewer (statistical) deaths in the least concentrated vs. mostconcentrated markets.

The English National Health Service (NHS) adopted a set of reforms in 2006 that wereintended to increase patient choice and hospital competition, and introduced regulatedprices for hospitals based on patient diagnoses (analogous to the Medicare ProspectivePayment System). Two recent studies examine the impacts of this reform (Cooper et al.,2010; Gaynor et al., 2010) and find that, following the reform, risk-adjusted mortalityfrom heart attacks fell more at hospitals in less concentrated markets than at hospitals inmore concentrated markets. Gaynor et al. (2010) also look at mortality from all causesand find that patients fared worse at hospitals in more consolidated markets.

The results of studies which examine impacts of competition for privately insured

9

patients are more mixed. A number of studies find that quality is positively affected bycompetition, a number find that it is negatively affected by competition, and some find noeffect. As a consequence, there is no clear impact of the impact of hospital consolidationon quality for privately insured patients that can be ascertained from the current researchliterature.

A recent study by Cutler et al. (2010) examines not only the impacts of competitionon quality, but also impacts on costs. Cutler et al. use the repeal of entry restrictingregulation (hospital certificate of need regulation; CON) in Pennsylvania to examinethe effect of entry of hospitals into the CABG surgery market. They find that entryled to increased quality, but that the gains from reduced mortality due to entry areapproximately offset by the additional costs incurred by entering firms.

A recent paper by Romano and Balan (2011) attempts to directly assess the im-pacts of hospital mergers on quality. Romano and Balan study the impact on quality ofcare of a consummated merger between two hospitals in the Chicago suburbs (EvanstonNorthwestern Hospital and Highland Park Hospital). This merger was the subject of anantitrust suit by the Federal Trade Commission, and the authors provided evidence onthe case. They find no significant impact of the merger on many quality measures, butthere is a significant negative impact on some and a few with positive impacts. Theyestimate that the merger led to heart attack, pneumonia, and stroke mortality going upat Evanston Northwestern Hospital, although not at Highland Park. There was someimprovement in quality for some nursing-sensitive quality measures: the incidence ofdecubitis ulcers (bedsores) fell at both merged entities, as did infections at EvanstonNorthwestern. Conversely, the incidence of hip fractures rose at Evanston Northwestern.Last, they found increases in some measures of obstetric outcomes (birth trauma to thenewborn, obstetric trauma to the mother), and decreases in some other measures. Theyconclude that overall there is no reason to infer that the merger had salutary effects onquality.

Overall, the research evidence suggests that hospital consolidation can have a negativeimpact on quality in markets with regulated prices, like Medicare. However, the currentresearch evidence where prices are market determined (the privately insured) does notindicate a clear impact of consolidation on quality in those markets in general.

3.2 Physicians

As stated previously, there is relatively little research on competition in physician servicesmarkets, mainly due to lack of available data. There are some studies, however, whichshed some light on this topic.

Economics suggests that there is a minimum population necessary for a given area tosupport a physician of a particular specialty. Further, the smaller the total number ofphysicians in a given specialty, the larger the population necessary to support a physician.

10

If the total number of physicians expands, then competition will lead to lower incomes forphysicians in larger communities, and physicians will diffuse out to smaller areas, so theminimum population necessary to support a physician will fall. Rosenthal et al. (2005)test this hypothesis by examining communities in states in which the total number ofphysicians doubled from 1970 to 1999. They find that communities of all sizes gainedphysicians over this period, but that the impact was larger for smaller communities,implying that competition among physicians did increase due to the increased numberof doctors.

Some recent work by Schneider et al. (2008) examines the impact of physician andinsurer market concentration on physician prices. They find that physician market con-centration is associated with significantly higher prices. A 1% increase in the physicianHHI leads to 1-4% higher physician prices. Conversely, they find that the health insurerHHI has no statistically significant impact on physician prices.

3.3 Insurers

Until recently there has been very little research on competition by health insurancefirms. Part of the reason for this has been the paucity of sufficient data for given marketsto construct measures of prices or market shares in the U.S. and internationally. Forthe U.S., there are detailed household level data (the Medical Expenditure Panel Survey,http://www.meps.ahrq.gov/mepsweb/) on health insurance decisions and prices, butthese data are from surveys of approximately 4,000 households selected to be nationallyrepresentative. As a consequence, they don’t fully describe the choices in a given market,or catalog the prices of alternatives, let alone the market shares of sellers. Other data(National Association of Insurance Commissioners, http://www.naic.org) do capturemarket shares, but at the level of a state. Since the vast majority of health insurancerestricts enrollees’ choices to a network of providers, most of whom are local, the geo-graphic market for health insurance is often local, and smaller than a state. In addition,there are no comprehensive data on health plans’ networks of providers. The most com-prehensive data on health insurance markets has been for private insurance associatedwith the Medicare program (Medigap coverage, Medicare+Choice/Medicare Advantage).

Dafny (2010) is one of the first of some new studies on insurance market competition.Dafny uses data from a benefits consulting firm on the plans purchased and premiumspaid by a large number (776) of large employers over the period 1998-2005. Whilethese data are not necessarily complete by market or nationally representative, they dorepresent the most extensive and comprehensive data set with prices and quantities forthe insurance market.

She examines the effect of shocks to employer profitability on the changes in theinsurance premiums they pay. The idea is that if insurers possess no market powerthen the premiums they charge will not vary with employer profitability. Only if insurers

11

have market power will they be able to price discriminate based on employer profitability.Dafny finds evidence that premiums increase with the buyer’s profitability. She also findsthat the effect of employer profitability on insurance premiums falls with the number offirms in the market. More precisely, what Dafny finds is that the effect of increasedprofitability for an employer is to increase premiums in markets with few insurers bymore than in markets with a large number of insurers, for the same employer.

It is notable that there are significant effects of employer profitability in insurancepremiums even for markets with 9-10 insurers (the effect is insignificant for markets with10 or more insurers). It seems surprising that insurers would possess market power inmarkets with 8-9 rivals (9-10 firms total). This empirical result does raise some doubts,but research by Dranove et al. (2003) found patterns in 1997 suggesting there may bemarket power in HMO markets with up to 6 firms. In any event, Dafny is a contributionthat opens a line of empirical research on competition in health insurance markets.

Dafny et al. (2011b) employ the same dataset as in Dafny (2010), but examine how thegrowth rate of an employer’s health insurance premiums is affected by health insurancemarket concentration (HHI). Initial estimates reveal no significant effect of insurer marketconcentration on premium growth. Of course, market concentration may be codeterminedwith premium growth, making it difficult to tease out a clear effect. In order to deal withthat problem Dafny et al. examine changes in local market concentration due to a largemerger in 1999 between two national health insurers: Aetna and Prudential Healthcare.Using this, they find a significant impact of the predicted change in HHI due to themerger on the change in premiums. They find that the cumulative effect of insurermarket consolidation on premiums is approximately 7%.

Dafny et al. also recognize that insurers may have bargaining power (market powerin a posted price world) with regard to providers. They therefore examine the effectof insurer concentration on changes in earnings and employment for physicians and fornurses as a way of testing for the presence of insurer monopsony power. They find thatthe merger reduced physician earnings growth on average by 3%, while nurses’ earningsrose by approximately 6-10ths of 1%. There is no significant effect on physician em-ployment, while nurse employment grows as a result of the merger induced increase inconcentration. This doesn’t seem to be evidence of monopsony power, but it is consis-tent with plan concentration leading to downward pressure on physician earnings andultimately to substitution of nurses for physicians.4 Overall, while the estimation resultsdepend crucially on the use of the Aetna-Prudential merger as a driver of changes inlocal insurance market concentration, they show evidence of a significant relationshipbetween changes in insurer market concentration and changes in premiums, implying alink between market structure and the exercise of market power.

4Sorensen (2003) and Wu (2009) find that health plans who are better able to channel patients canextract greater discounts from hospitals. Shen et al. (2010) and Moriya et al. (2010) find health insurerconcentration reduces hospital prices, while Schneider et al. (2008) find no effect of insurer concentrationon physician prices.

12

Maestas et al. (2009) document substantial price variation in Medigap insurance mar-kets, even though plans are standardized. They find that insurers have substantial dif-ferences in costs, and hence loading fees, which contribute to the observed variation inprices. This conclude that price dispersion is caused by substantial search costs – Maes-tas et al. estimate an average search cost for consumers in the market of $72 and amaximum of $144.

Starc (2010) also examines the Medigap market for insurance, using a model thatallows for both adverse selection and market power. She documents that the marketis highly concentrated – the national 4 firm concentration ratio is 83% (compared to44% for private passenger automobile insurance or 34% for life insurance), and two firms(UnitedHealth, 46% and Mutual of Omaha, 24%) account for almost all of that. Starcdocuments substantial price dispersion for Medigap policies, confirming Maestas et al.,and documents a positive relationship between premiums and market concentration, thesame qualitative result as in Dafny et al. (2011b), albeit for a very different market.A 1% increase in the 2 firm concentration ratio is associated with a 0.26% increase inpremiums.

Starc then uses the estimates from her model to find the impacts of adverse selectionand market power in the Medigap market. She finds that adverse selection increasespremiums by 9% and reduces the size of the market (lowers insurance coverage) by 18%.Starc goes on to calculate the impact of market power by calculating what premiumswould be if they were set equal to costs. This is estimated to lead to large reductions inprices – 44 and 45%, respectively. These results emphasize the substantial market powerexercised in this market.

Lustig (2010) examines the market for Medicare+Choice plans in 2000-2003. Medi-care+Choice are private managed care plans that Medicare beneficiaries may choose asan alternative to traditional Medicare (the current version of this program is called Medi-care Advantage). Lustig allows for both adverse selection and market power in his model.He finds that consumers’ health risk has no significant impact on insurers’ costs. Lustigthen goes on to use the model estimates to simulate the gains to consumers and producerswhen adverse selection is eliminated and compare that to what he actually observes. Thegain from eliminating adverse selection is simulated for markets with increasing numbersof insurers to generate effects of competition. Lustig finds that the gains to eliminat-ing adverse selection increase steadily in the number of insurance firms. For example,in one of his simulations Lustig finds that removing adverse selection eliminates 17%of the difference in consumer and producer benefits between the observed and sociallyoptimal outcomes where there’s a monopoly, while it eliminates 35% of the difference ina duopoly, and 50% where there are 6 or more firms. This implies that when there ismarket power most of the welfare loss is due to the exercise of market power as opposedto adverse selection.

A paper by Town and Liu (2003) estimates a model of the Medicare+Choice market,

13

the predecessor of Medicare Advantage, i.e., the market for private Medicare plans. Theyuse their estimates to calculate the impacts of the program and of competition. Theyfind that the creation of the M+C program resulted in approximately $15.6 billion inconsumer surplus and $52 billion in profits from 1993 to 2000 (in 2000 dollars), i.e., $67.6billion in value to consumers and producers. They also find evidence of competitiveeffects. Consumer benefit increases with the number of plans in a county, and most of theincrease is due to increased premium competition. Comparing monopoly markets versusmarkets with 4 firms, they find that 81% of the difference in consumer benefit (higher inquadropoly markets) is due to increased premium competition. Of the remainder, 3% isfrom increased product variety and 8% from prescription drug coverage.

3.4 Vertical Integration

Vertical integration between hospitals and physicians or insurers and providers can inprinciple provide efficiencies by aligning incentives, allowing for better coordination ofcare and joint investments which enhance efficiency or the quality of care. At the sametime, integration can potentially harm competition by foreclosing rivals from access tokey inputs. An integrated system which has locked up all the orthopedists in town, forexample, may make it difficult to impossible for another hospital to offer orthopedic ser-vices or for a freestanding ambulatory surgery center to enter the market and compete onorthopedic services. Separately, integration may eliminate competition among previouslyindependent providers. For example, physicians who had previously been in competitionall become members of the same firm once they integrate with a hospital system (or aninsurer).

There is very little evidence at present on the impact of vertical integration on marketpower. In part, that is because vertical integration has not been that common in healthcare. It was quite rare until the mid-1990s, and then declined rapidly thereafter. Inte-gration between hospitals and physician practices peaked in 1996 at approximately 40%of all hospitals, and declined thereafter (Burns and Pauly, 2002; Ciliberto, 2005). Thispattern was repeated with vertical integration of hospitals into the insurance market,although the extent of vertical integration was never as great as between hospitals andphysicians (Burns and Pauly, 2002). This growth coincided with the growth of managedcare, and in particular with the perceived growth in managed care organizations’ nego-tiating power with hospitals. Burns et al. (2000) find that hospital-physician alliancesincrease with the number of HMOs in the market. They infer that providers may beintegrating in order to achieve or enhance market power. More recently, Berenson et al.(2010) conducted 300 interviews with health care market participants, and report thatincreased bargaining power through joint negotiations listed as one of several reasons forhospital-physician alliances.

Certain types of vertical relations in health care have been the subject of significantantitrust scrutiny — exclusive dealing between physician practices and hospitals (usually

14

for a specialized service, e.g., radiology, anesthesiology, or pathology), and most-favored-nations clauses between insurers and providers, which require the provider to give theinsurer a rate as low as it gives to any buyer (see Gaynor and Haas-Wilson, 1998; Haas-Wilson, 2003, for reviews of vertical issues in health care).5

In spite of the interest in this topic, there is relatively little evidence on the effects ofvertical restraints in health care. Ciliberto and Dranove (2005) and Cuellar and Gertler(2005) are the only two papers (of which I am aware) which examine the competitiveimpacts of vertical integration in health care. Both papers look at the effects of hospital-physician practice integration on hospital prices. The two studies find opposite results– Cuellar and Gertler find evidence consistent with anticompetitive effects of physician-hospital integration, while Ciliberto and Dranove find no such evidence.

Research on efficiencies from integration does not find much evidence of positive gainsfrom integration. Burns and Muller (2008) review the empirical evidence on hospital-physician relationships. They find little evidence of an impact of integration on costs,quality, access or clinical integration. Madison (2004) investigates the relationship be-tween hospital-physician affiliations and patient treatments, expenditures, and outcomesusing data on Medicare heart attack patients. She finds little evidence of any impacts ofhospital-physician relationships.

4 Who Pays for Higher Provider Prices?

As mentioned previously, when consolidation leads to providers obtaining higher pricesfrom providers the impact ultimately falls on consumers, not insurers or employers. Re-search evidence consistently shows that higher health insurance costs for employers arepassed on to employees in the form of lower wages (or lower wage increases) (see, e.g.,Gruber, 1994; Bhattacharya and Bundorf, 2005; Baicker and Chandra, 2006; Emanueland Fuchs, 2008). In addition, there is recent evidence that hospital mergers have in-creased the number of the uninsured. Town et al. (2006) find that private insurance rollsdeclined by approximately 0.3 percentage points or approximately 695,000 lives in 2003due to the effects of hospital mergers, with the vast majority of those who lost privateinsurance joining the ranks of the uninsured.

As a consequence, even though privately insured consumers typically pay a verysmall share of provider fees directly, they bear the brunt of the impact of higher fees by

5Complaints about exclusive contracts between hospitals and physician practices are the most nu-merous type of antitrust case brought in health care. One of these cases was decided by the SupremeCourt (Jefferson Parish Hosp. Dist. No. 2 v. Hyde, 466 U.S. 2(1984)), and represents an importantlegal precedent on exclusive dealing and tying. There have also been a number of cases on most-favored-nations clauses, e.g., Ocean State Physicians Health Plan v. Blue Cross & Blue Shield, 883 F.2d 1101(1st Cir. 1989), cert. denied, 494 U.S. 1027 (1990) and Blue Cross & Blue Shield v. Marshfield Clinic,65 F.3d 1406 (7th Cir. 1995), cert. denied, 116 S. Ct. 1288 (1996).

15

ultimately receiving reduced total compensation from their employers.

5 Policy

There are a number of policy levers that can be considered to deal with consolidation inthe health care industry. They include antitrust enforcement, enhancing supply by facil-itating the entry of new providers (including new forms of providers), policies designedto enhance the responsiveness of demand, and rate regulation. These policies need notbe pursued in isolation. For example, vigorous antitrust enforcement is likely to be mosteffective paired with policies that facilitate entry. Even under rate regulation providerscompete for patients via quality or other means. Taking steps to maintain or enhancethat competitive environment can substantially improve performance even if prices areregulated. Policies that enhance the entry of new providers can ultimately reduce theneed for intensive antitrust enforcement or rate regulation.

5.1 Antitrust

The standard approach to the injurious effects of market consolidation is vigorous an-titrust enforcement. Antitrust enforcement must be applied judiciously. Consolidationor integration can be efficiency enhancing, and those cases should be allowed and en-couraged. However, the evidence does not support efficiency enhancing consolidation ingeneral, at least in hospital markets. While there is some evidence that prices are higherin more concentrated physician and insurance markets, we do not have direct researchevidence on the impacts of consolidation on efficiencies or quality in those markets.6

Antitrust enforcers should carefully examine proposed mergers and block those theyfind anticompetitive. This policy can be successfully pursued, but it should be recognizedthat many markets (especially for hospitals) are already highly concentrated. Anotherpolicy is to prosecute cases against consummated mergers, where the merged entitieshave exercised market power and not realized any substantial efficiencies. In addition,enforcers can pursue anticompetitive behavior intended to harm competition (for exam-ple, attempting to prevent entry or force exit of competitors). The enforcement agencies’resources are limited, so there are only so many cases of this sort they can pursue. Itis possible, however, that vigorous enforcement against egregious offenders may havea chilling effect on the exercise of market power by others, who fear the possibility ofdetection and prosecution.

Another aspect of antitrust enforcement is providing safe harbors for integration thatis considered to be efficiency enhancing. While efficient forms of organization should be

6Nonetheless, if prices are higher in more concentrated markets, that is evidence that any efficienciesthat consolidated firms may realize are not being passed on to consumers in the form of lower prices.

16

allowed to emerge and flourish, care must also be taken to avoid allowing integration thatis only a sham with regard to efficiency and is truly for the purpose of enhancing marketpower. This is true in general, but is currently salient with regard to the Accountable CareOrganizations (ACOs) which are part of the health reform law. As indicated previously,integration among hospitals and doctors has the potential to be efficiency enhancing,but can also cause harm to competition, and can potentially make it more difficult forinnovative organizations to enter markets. This is a factor which must be given seriousconsideration.

Last, it is sometimes proposed that allowing firms on one side of a market to acquiremarket power is an appropriate response if firms on the other side of the market alreadypossess it (e.g., allowing providers to acquire market power if insurers have substantialmarket power). This is sometimes called countervailing power. The notion is that havinga sort of balance of market power on both sides of the market is better than only havingmarket power on one side. This is not necessarily the case. First, it must be true thatone side of the market does indeed have market power and it is causing harm. Hardbargaining over price (or other matters) is not in and of itself of harmful market power.Second, the best response is to deal directly with the problem, and try to eliminate orreduce the extant market power. Antitrust enforcement is the most likely tool for this.This is certainly better than creating a situation with more market power. Third, if it isnot possible for some reason to deal directly with the extant market power on one sideof the market, having market power on both sides does not necessarily improve matters.It can do so, but it can also make things worse. If the only way firms have to increasetheir bargaining power is by restricting the quantity they sell, then countervailing powerwill make things worse than market power on only one side of the market.

5.2 Supply Side Policies

Another policy lever are policies that facilitate the entry of new providers, and in partic-ular allow for innovation. It may be that markets (again, hospital markets in particular)have become so concentrated that the effectiveness of antitrust enforcement is limited.That may leave inefficiencies in place in the short run. However, in the long run, as thetechnology of medicine changes, new types of providers will likely emerge and competewith current organizational forms. We have already seen some of this via the entry offree standing ambulatory surgery centers, retail clinics, and specialty hospitals. In thecase of insurance, a key factor governing entry is the ability to field an attractive net-work of providers. Making sure that providers, particularly those that are perceived as“must-have” providers, are available to potential entrants is a critical factor. Regulationcan play a role as well. Relatively few states have retained certificate of need regulationfor health care facilities, but such regulation can raise the costs of entry. Insurers haveto deal with differing sets of regulations in every state, and are not allowed to pool riskacross states. Such factors may also raise the costs of entry.

17

Facilitating the entry of new firms, and in particular innovative new forms, can amelio-rate or eliminate the effects of current consolidation in the long run. As stated previously,however, care still must be taken to ensure that such entry is possible, and not foreclosedby integration, so antitrust enforcement also plays a key role here. Obviously, such aprocess will take time, although change can sometimes occur very rapidly.

5.3 Demand Side Policies

Competition can only occur if demand by payers or consumers is sufficiently responsive toprice or quality differences across providers, and if there’s sufficient choice of providers.Selective contracting by private payers is a mechanism which helps to create an envi-ronment where competition among providers is possible. If payers contract with everyprovider in a market, then they have very little bargaining power. Policies that enhancethe ability of payers to selectively contract with providers are important, but they willonly be effective if there is sufficient choice among providers.

Transparency, providing information about prices or quality to the public, is a policythat has received substantial attention, and has been implemented by some states. Inprinciple, it seems as more information should be better. However, this is not necessarilythe case. Making pricing information public can make it easier for firms to collude.In addition, it’s not clear that heavily insured consumers have sufficient incentive topay attention to price differences. Even consumers who have policies with a lot of costsharing will not face much of the impacts of price differences if they obtain an expensivetreatment. Expensive treatments put most consumers well beyond their deductibles andcopays so that they bear little to none of any price differences across providers.7 Ofcourse, expensive treatments account for the majority of medical spending.

Providing clear and understandable information about products (providers’ and in-surers’) so consumers can understand what they are obtaining can facilitate competition(again, conditional on sufficient alternatives). If consumers have little information ordon’t understand the information they have, they tend to rely on reputations, brandnames, etc. This tends to decrease the responsiveness of demand to prices or otherfactors and enhances firms’ market power.

5.4 Rate Regulation

A possible response to highly concentrated markets is rate regulation. If markets arehighly concentrated and there appears little hope of competition, then price regulationis an economic policy that can be employed. However, moving to a regulated price

7This doesn’t mean that being insured against large losses is bad – it isn’t. Consumers should beinsured against large risks. It just means that it’s not realistic to expect them to pay attention to pricesin such a situation.

18

environment for the privately insured would be a nontrivial undertaking, with all of theissues associated with regulation and bureaucracy. Evidence from state hospital rateregulation programs in the 1970s and 1980s indicates a mixed pattern of success. Thesetup and administration of the program played a large role in whether they were effective.Nonetheless, there is evidence that finds that mandatory rate regulation program in anumber of states did reduce the rate of growth of hospital expenses (by a little more than1%) (Dranove and Cone, 1985; Sloan, 1983).

19

Bibliography

Akosa Antwi, Y. O. D., Gaynor, M., and Vogt, W. B. (2009). A bargain at twice theprice? California hospital prices in the new millennium. Forum for Health Economicsand Policy, 12(1):Article 3. http://www.bepress.com/fhep/12/1/3.

American Medical Association (2010). Competition in health insurance: A comprehensivestudy of U.S. markets. Technical report, American Medical Association, Chicago, IL.2006-2010 editions.

Baicker, K. and Chandra, A. (2006). The labor market effects of rising health insurancepremiums. Journal of Labor Economics, 24(3):609–634.

Berenson, R., Ginsburg, P., and Kemper, N. (2010). Unchecked provider clout in Cali-fornia foreshadows challenges to health reform. Health Affairs, 29(4):699.

Bhattacharya, J. and Bundorf, M. K. (2005). The incidence of the healthcare costs ofobesity. National Bureau of Economic Research Working Paper No. 11303.

Brand, K., Garmon, C., Gowrisankaran, G., Nevo, A., and Town, R. (2011). Estimatingthe price impact of hospital mergers: Inova’s proposed acquisition of Prince Williamhospital. unpublished manuscript, Federal Trade Commission, University of Arizona,Northwestern University, University of Minnesota.

Burns, L., Bazzoli, G., Dynan, L., and Wholey, D. (2000). Impact of HMO marketstructure physician-hospital strategic alliances. Health Services Research, 35(1):101–132.

Burns, L. and Muller, R. (2008). Hospital-physician collaboration: Landscape of eco-nomic integration and impact on clinical integration. Milbank Quarterly, 86(3):375–434.

Burns, L. R. and Pauly, M. V. (2002). Integrated delivery networks: A detour on theroad to integrated health care? Health Affairs, 21(4):128–143.

Capps, C., Carlton, D., and David, G. (2010). Antitrust treatment of nonprofits: Shouldhospitals receive special care? unpublished manuscript, Bates White LLC, Universityof Chicago, University of Pennsylvania.

Capps, C. and Dranove, D. (2004). Hospital consolidation and negotiated PPO prices.Health Affairs, 23(2):175–181.

Capps, C., Dranove, D., and Satterthwaite, M. (2003). Competition and market powerin option demand markets. RAND Journal of Economics, 34(4):737–63.

20

Capps, C. S. (2009). Federal health plan merger enforcement is consistent and robust.White Paper, Bates White LLC.

Chernew, M. (1995). The impact of non-IPA HMOs on the number of hospitals andhospital capacity. Inquiry, 32(2):143–154.

Ciliberto, F. (2005). Does organizational form affect investment decisions? unpublishedmanuscript, University of Virginia.

Ciliberto, F. and Dranove, D. (2005). The effect of physician-hospital affiliations onhospital prices in California. Journal of Health Economics, 25(1):29–38.

Cooper, Z., Gibbons, S., Jones, S., and McGuire, A. (2010). Does hospital competi-tion save lives? Evidence from the English NHS patient choice reforms. unpublishedmanuscript, London School of Economics.

Cuellar, A. E. and Gertler, P. J. (2005). Strategic integration of hospitals and physicians.Journal of Health Economics, 25(1):1–28.

Cutler, D., McClellan, M., and Newhouse, J. (2000). How does managed care do it?RAND Journal of Economics, 31(3):526–548.

Cutler, D. M., Huckman, R. S., and Kolstad, J. T. (2010). Input constraints and the effi-ciency of entry: Lessons from cardiac surgery. American Economic Journal: EconomicPolicy, 2(1):51–76.

Dafny, L. (2009). Estimation and identification of merger effects: An application tohospital mergers. Journal of Law and Economics, 52(3):pp. 523–550.

Dafny, L. (2010). Are health insurance markets competitive? American EconomicReview, 100:1399–1431.

Dafny, L., Dranove, D., Limbrock, F., and Scott Morton, F. (2011a). Data impedimentsto empirical work in health insurance markets. unpublished manuscript, NorthwesternUniversity, Yale University.

Dafny, L., Duggan, M., and Ramanarayanan, S. (2011b). Paying a premium on yourpremium? Consolidation in the U.S. health insurance industry. American EconomicReview, forthcoming.

Dranove, D. and Cone, K. (1985). Do state rate setting regulations really lower hospitalexpenses? Journal of Health Economics, 4:159–165.

Dranove, D., Gron, A., and Mazzeo, M. (2003). Differentiation and competition in HMOmarkets. Journal of Industrial Economics, 51(4):433–454.

21

Dranove, D. and Lindrooth, R. (2003). Hospital consolidation and costs: Another lookat the evidence. Journal of Health Economics, 22:983–997.

Dranove, D. and Ludwick, R. (1999). Competition and pricing by nonprofit hospitals: areassesment of Lynk’s analysis. Journal of Health Economics, 18:87–98.

Dranove, D., Simon, C., and White, W. (2002). Is managed care leading to consolidationin health-care markets? Health Services Research, 37(3):573–594.

Dranove, D. D. and Satterthwaite, M. A. (2000). The industrial organization of healthcare markets. In Culyer, A. and Newhouse, J., editors, Handbook of Health Economics,chapter 20, pages 1094–1139. Elsevier Science, North-Holland, New York and Oxford.

Emanuel, E. and Fuchs, V. R. (2008). Who really pays for health care? The myth of“shared responsibility”. Journal of the American Medical Association, 299(9):1057–1059.

Federal Trade Commission and Department of Justice (1992). Horizontal merger guide-lines. Issued April 2, 1992, Revised September, 2010.

Fuchs, V. (2007). Managed care and merger mania. Journal of the American MedicalAssociation, 277(11):920–921.

Gaynor, M. (2006). What do we know about competition and quality in health care mar-kets? Foundations and Trends in Microeconomics, 2(6). http://www.nowpublishers.com/product.aspx?product=MIC\&doi=0700000024.

Gaynor, M. and Haas-Wilson, D. (1998). Vertical relations in health care markets. InMorrisey, M. A., editor, Managed Care and Changing Health Care Markets, chapter 7,pages 140–163. AEI Press, Washington, DC.

Gaynor, M., Moreno-Serra, R., and Propper, C. (2010). Death by market power: Reform,competition and patient outcomes in the British National Health Service. unpublishedmanuscript, Carnegie Mellon University, Imperial College.

Gaynor, M. and Town, R. J. (2012). Competition in health care markets. In McGuire,T. G., Pauly, M. V., and Pita Barros, P., editors, Handbook of Health Economics,volume 2, chapter 9. Elsevier North-Holland, Amsterdam and London. forthcoming.

Gaynor, M. and Vogt, W. B. (2000). Antitrust and competition in health care markets.In Culyer, A. and Newhouse, J., editors, Handbook of Health Economics, chapter 27,pages 1405–1487. Elsevier Science, North-Holland, New York and Oxford.

Gaynor, M. and Vogt, W. B. (2003). Competition among hospitals. Rand Journal ofEconomics, 34(4):764–785.

22

Ginsburg, P. (2010). Wide variation in hospital and physician payment rates evidenceof provider market power. HSC Research Brief No. 16, Center for Studying HealthSystem Change.

Government Accountability Office (2009). State small group health insurance markets.Technical report, U.S. Government Accountability Office, Washington, DC. GAO-09-363R.

Gruber, J. (1994). The incidence of mandated maternity benefits. American EconomicReview, 84:622–641.

Haas-Wilson, D. (2003). Managed Care and Monopoly Power: The Antitrust Challenge.Harvard University Press, Cambridge, MA.

Haas-Wilson, D. and Garmon, C. (2011). Hospital mergers and competitive effects: Tworetrospective analyses,. International Journal of the Economics of Business, 18(1):17–32.

Harrison, T. (2010). Do mergers really reduce costs? Evidence from hospitals. EconomicInquiry, 49:1–16.

Keeler, E., Melnick, G., and Zwanziger, J. (1999). The changing effects of competitionon non-profit and for-profit hospital pricing behavior. Journal of Health Economics,18:69–86.

Kessler, D. and McClellan, M. (2000). Is hospital competition socially wasteful? Quar-terly Journal of Economics, 115(2):577–615.

Kessler, D. P. and Geppert, J. J. (2005). The effects of competition on variation in thequality and cost of medical care. Journal of Economics and Management Strategy,14(3):575–589.

Krishnan, R. (2001). Market restructuring and pricing in the hospital industry. Journalof Health Economics, 20:213–237.

Liebhaber, A. and Grossman, J. M. (2007). Physicians moving to mid-sized, single-specialty practices. Technical report, Center for Studying Health System Change,Washington, DC. Tracking Report No. 18.

Lustig, J. (2010). Measuring welfare losses from adverse selection and imperfect compe-tition in privatized Medicare. unpublished manuscript, Boston University.

Madison, K. (2004). Hospital–physician affiliations and patient treatments, expenditures,and outcomes. Health Services Research, 39(2):257–278.

23

Maestas, N., Schroeder, M., and Goldman, D. (2009). Price variation in markets withhomogeneous goods: The case of Medigap. Working Paper No. 14679, National Bureauof Economic Research.

Martin, A., Lassman, D., Whittle, L., and Catlin, A. (2011). Recession contributes toslowest annual rate of increase in health spending in five decades. Health Affairs,30:111–122.

Massachusetts Attorney General (2010). Investigation of health care cost trends and costdrivers, pursuant to G.L. c.118G, § 6 1/2(b). Preliminary report, Office of AttorneyGeneral Martha Coakley, Boston, MA. January 29.

Moriya, A. S., Vogt, W. B., and Gaynor, M. (2010). Hospital prices and market structurein the hospital and insurance industries. Health Economics, Policy and Law, 5:459–479.

National Center for Health Statistics (2011). Health, United States, 2010. Technicalreport, National Center for Health Statistics, Center for Disease Control, Hyattsville,MD.

Pennsylvania Health Care Cost Containment Council (2007). Cardiac surgery in Pennsyl-vania 2005. Report, Pennsylvania Health Care Cost Containment Council, Harrisburg,PA. June.

Romano, P. and Balan, D. (2011). A retrospective analysis of the clinical quality effectsof the acquisition of Highland Park hospital by Evanston Northwestern healthcare.International Journal of the Economics of Business, 18(1):45–64.

Rosenthal, M. B., Zaslavsky, A., and Newhouse, J. P. (2005). The geographic distributionof physicians revisited. Health Services Research, 40(6, Part 1):1931–1952.

Sacher, S. and Vita, M. (2001). The competitive effects of a not-for-profit hospital merger:a case study. Journal of Industrial Economics, 49(1):63–84.

Schneider, J., Li, P., Klepser, D., Peterson, N., Brown, T., and Scheffler, R. (2008). Theeffect of physician and health plan market concentration on prices in commercial healthinsurance markets. International Journal of Health Care Finance and Economics, 8:13–26.

Shen, Y., Wu, V., and Melnick, G. (2010). Trends in hospital cost and revenue, 1994-2005:How are they related to HMO penetration, concentration, and for-profit ownership?Health Services Research, 45(1):42–61.

Simpson, J. and Shin, R. (1997). Do non-profit hospitals exercise market power? un-published manuscript, Federal Trade Commission.

Sloan, F. A. (1983). Rate regulation as a strategy for hospital cost control: Evidence fromthe last decade. Milbank Memorial Fund Quarterly/Health and Society, 61:195–221.

24

Sorensen, A. T. (2003). Insurer-hospital bargaining: Negotiated discounts in post-deregulation Connecticut. Journal of Industrial Economics, 51(4):469–490.

Spang, H., Bazzoli, G., and Arnould, R. (2001). Hospital mergers and savings for con-sumers: Exploring new evidence. Health Affairs, 20(4):150–158.

Starc, A. (2010). Insurer pricing and consumer welfare: Evidence from Medigap. unpub-lished manuscript, Harvard University.

Tenn, S. (2011). The price effects of hospital mergers: A case study of the sutter-summittransaction. International Journal of the Economics of Business, 18(1):65–82.

Thompson, E. (2011). The effect of hospital mergers on inpatient prices: A case studyof the New Hanover-Cape Fear transaction,. International Journal of the Economicsof Business, 18(1):91–101.

Town, R. and Vistnes, G. (2001). Hospital competition in HMO networks. Journal ofHealth Economics, 20(5):733–752.

Town, R., Wholey, D., Feldman, R., and Burns, L. (2006). The welfare consequences ofhospital mergers. Working Paper No. 12244, National Bureau of Economic Research.

Town, R., Wholey, D., Feldman, R., and Burns, L. (2007). Revisiting the relationshipbetween managed care and hospital consolidation. Health Services Research, 42(1):219–238.

Town, R. J. and Liu, S. (2003). The welfare impact of Medicare HMOs. RAND Journalof Economics, 34:719–736.

Town, R. J. and Park, M. (2011). Market structure beliefs and hospital merger waves.unpublished manuscript, University of Minnesota.

Vogt, W. B. and Town, R. J. (2006). How has hospital consolidation affected the priceand quality of hospital care? Research Synthesis Report 9, Robert Wood JohnsonFoundation, Princeton, NJ.

Wu, V. (2009). Managed care’s price bargaining with hospitals. Journal of Health Eco-nomics, 28:350–360.

25

Table 1: Hospital Market Concentration,U.S., 1987-2006a

Year Mean HHIb Changec

1987 2,340 —1992 2,440 1001997 2,983 5432002 3,236 2532006 3,261 25a Source: American Hospital Association.

Data are for U.S. Metropolitan Statisti-cal Areas with population < 3 million.

b Herfindahl-Hirschmann Index. Meansweighted by MSA population.

c Total change from the previous year inthe table.

26

Figure 1: Hospital Mergers and Acquisitions, 1998-2010. Source: Irving Levin Associates,Inc., The Health Care Acquisition Report, Sixteenth Edition, 2010.

287

249

300

350 Number of Deals

Number of Hospitals

139175

132118

236 249

149112

150

200

250

11086 83

5838

59 51 57 58 60 5275

56

118101

56

88 78 80

112

0

50

100

01998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011*

Figure 2: Number of Hospitals in Health Systems, 2000-2009. Source: American HospitalAssociation.

3,000

2,800

2,900

s

2 600

2,700

Hos

pita

ls

2,500

2,600

2,4002000 2001 2002 2003 2004 2005 2006 2007 2008 2009

27

Figure 3: Scatterplot of MSA HHI in 1990 and HHI in 20060

0

02000

2000

20004000

4000

40006000

6000

60008000

8000

800010000

1000

010000HHI in 2006HH

I in

2006

HHI in 20060

0

02000

2000

20004000

4000

40006000

6000

60008000

8000

800010000

10000

10000HHI in 1990

HHI in 1990

HHI in 1990

Figure 4: Trends in Hospital Concentration, M&A Activity and HMO Penetration: 1990-2006

0 

10 

20 

30 

40 

50 

60 

70 

80 

90 

0 

0.05 

0.1 

0.15 

0.2 

0.25 

0.3 

0.35 

1988  1990  1992  1994  1996  1998  2000  2002  2004  2006  2008 

Num

ber of M

&As 

HHI / HMO Pen

etra6on

  

Year 

HHI  HMO Penetra7on  Number of MSA M&As 

28

Table 2: Insurance Market Concentration, U.S., 2004-2008a,b

Year Median HHI Change Mean HHI Change

2004 3,544 — 3,939 —2005 3,748 204 4,077 1382006 2,986 -762 3,440 -6372007 3,558 572 3,944 5042008 3,276 -282 3,727 -217a Source: See American Medical Association (2010) for

more information on the data and calculations. Amer-ican Medical Association (AMA) calculations for thecombined HMO+PPO markets using January 1st en-rollment data from HealthLeaders-InterStudy’s (HLIS)Managed Market Surveyors c© HealthLeaders- Inter-Study.

b MSA-level HHIs for HMO+PPO markets.

29

Table 3: Large Employer Insurance Market Concentration, U.S., 1998-2009a

Self + Fully Insured Fully Insured OnlyYear Mean HHIb Change Mean HHIb Change

1998 2,172 — 2,984 —1999 1,997 -175 2,8352000 2,175 178 3,0922001 2,093 -82 3,0062002 2,280 187 3,1582003 2,343 63 3,4322004 2,519 176 3,7062005 2,609 90 3,9512006 2,740 131 4,0722007 2,873 133 4,0562008 2,916 43 4,2012009 2,956 40 4,126a Source: These #s were graciously provided by Leemore Dafny. The data

are for large multisite employers and do not represent the totality of theinsurance market. For more information on the data source, see Dafny(2010).

b Weighted by # of enrollees.

Table 4: Small Group Insurance Market Structure, U.S.,2000-2003a

Year Median Market Share, # of States with 5 firmLargest Carrier concentration ratio ≥ 75%

2002 33% 19 (of 34; 56%)2005 43% 26 (of 34; 77%)2008 47% 34 (of 39; 87%)a Source: Government Accountability Office (2009).

30

Figure 5: Physician-Hospital Trends, 1994-2007. Source: American Hospital Association.

30%

35%

20%

25%

30%

Hosp

itals

EmploymentPhysician Hospital Organization

10%

15%

20%

cent

age

of

IPA

Management Service Organization

0%

5%

10%

Perc

Group Practice without Walls

0%94 95 97 98 99 00 01 02 03 04 05 06 07

31